Activated carbon induced oxygen vacancies-engineered nickel ferrite with enhanced conductivity for supercapacitor application

Abstract

NiFe₂O₄ is a kind of bimetallic oxide possessing excellent theoretical capacity and application prospect in the field of supercapacitors. Whereas, due to the inherent poor conductivity of metal oxides, the performance of NiFe₂O₄ is not ideal in practice. Oxygen vacancies can not only enhance the conductivities of NiFe₂O₄ but also provide better adsorption of OH, which is beneficial to the electrochemical performances. Hence, oxygen vacancies engineered NiFe₂O₄ (NiFe₂O_4‒δ) is obtained through a two-step method, including a hydrothermal reaction and a further heat treatment in activated carbon bed. Results of electron paramagnetic resonance spectra indicate that more oxygen vacancies exist in the treated NiFe₂O_4‒δ than the original one. UV-Vis diffuse reflectance spectra prove that the treated NiFe₂O_4‒δ owns better conductivity than the original NiFe₂O₄. As for the electrochemical performances, the treated NiFe₂O_4‒δ performs a high specific capacitance of 808.02 F∙g^‒1 at 1 A∙g^‒1. Moreover, the asymmetric supercapacitor of NiFe₂O_4‒δ//active carbon displays a high energy density of 17.7 Wh∙kg^‒1 at the power density of 375 W∙kg^‒1. This work gives an effective way to improve the conductivity of metal oxides, which is beneficial to the application of metal oxides in supercapacitors.

Key words： nickel ferrite; oxygen vacancy; high conductivity; supercapacitor

1 Introduction

With the evolution of electronic industry, portable energy storage devices with high efficiency have developed into a promising direction in the energy field [1]. Among all the energy storage devices, researchers attach more and more importance to supercapacitors due to their excellent performances of rapid charge–discharge rate, high power density and stable cyclic charge–discharge ability [2]. In general, two kinds of energy storage mechanism exist in the storage process of supercapacitors. One is regarded as double-layer storage mechanism, which utilizes the charge accretion in the interface of electrode and electrolyte to storage energy. The other one is considered to be pseudocapacitance storage mechanism, relying on the reversible chemical adsorption–desorption or redox reaction to reserve energy [3].

As the pivotal component of supercapacitors, electrode materials have been invested a large amount of vigor. Typically, electrode materials can be divided into carbon based materials [4,5], metal oxides [6,7] and conducting polymer [8,9]. Thereinto, metal oxides possess broad potential application prospects due to their high theoretical energy density, while the practical applications are limited by their low conductivity. Zhou et al. [10] prepared ultra-small Fe₂O₃ nanodots skillfully integrated on reduced graphene oxide (rGO) sheets by a deposition method. The specific capacitance could arrive at 734.2 C∙g^‒1 at the current density of 2 A∙g^‒1 and maintain 77.9% at 20 A∙g^‒1. Meanwhile, they also proved that without the addition of rGO, the specific capacitance of Fe₂O₃ nanoparticles could only retain 15.8% at 20 A∙g^‒1. In another research, interconnected NiS/Co₃S₄ nanosheets grown on Fe₂O₃ substrate were synthesized by Guo et al. [11] through a facile in situ solution vulcanization process. The obtained materials exhibited awesome specific capacitance of 1213.7 F∙g^‒1 at the current density of 1 A∙g^‒1, while it decreased rapidly to 785.5 F∙g^‒1 at the current density of 10 A∙g^‒1. Only 64.7% capacity could retain. The low conductivity has been demonstrated to be the main factor to restrict the practical utilization of metal oxides.

Comparatively speaking, bimetallic spinel structure oxides perform a decent electronic conductivity compared with monometallic oxides due to the existence of abundant metal ions with different valences. Meanwhile, the synergistic effect between different metal ions leads to a higher energy capacity. Therein, NiFe₂O₄ attracts lots of researchers’ attention. NiFe₂O₄ is a kind of inverse spinel structure material, in which half of iron ions occupy the tetrahedral sites and nickel ions together with the other half of iron ions take up the octahedral sites. Redox pairs of Ni²⁺/Ni³⁺ and Fe²⁺/Fe³⁺ serve as active sites for rapid and reversible redox reactions [12]. Mordina et al. [13] designed an extremely stable hybrid supercapacitor employing binder free NiFe₂O₄ as the positive electrode and activated carbon as the negative electrode. An innovative surfactant-assisted co-precipitation method was used to prepare the Ni-foam/ NiFe₂O₄ positive electrode. Results showed that specific capacity of 398 C∙g^‒1 at 1 A∙g^‒1 was achieved, while it declined to 42 C∙g^‒1 at 20 A∙g^‒1. Poor conductivity is still the most significant restriction for NiFe₂O₄, and combining with carbon materials has been proved to be an efficient way to solve the problem. Yu et al. [14] fabricated NiFe₂O₄ nanoparticles coated on carbon cloth for stable and high-performance flexible supercapacitors, the specific capacitance of which could maintain 80% as the current density increasing from 2 to 100 mA∙cm^‒2.

In addition to the above-mentioned method, it is also important to improve NiFe₂O₄’s own electrical conductivity. As a typical spinel structure oxide, the conductive mechanism of NiFe₂O₄ can be explicated by small polaron-hopping model [15]. It has been proposed that the electron conduction of spinel is due to charge transfer between cations with different valences in octahedral sites [16]. Therefore, adjusting the ratio of Fe²⁺ and Fe³⁺ ions in the NiFe₂O₄ is considered to benefit the electrochemical properties, which can be achieved through forming abundant oxygen vacancy. Zhang et al. [17] produced MnO₂ with rich bulk oxygen vacancies by a complex induced chemical precipitation method. The obtained material performed enhanced charge transfer kinetics and was used for supercapacitor. A high specific capacitive retention of 78% was achieved with the increasing current density from 1 to 20 A∙g^‒1, exhibiting well-improved electrical conductivity. Besides, oxygen vacancies can not only improve the conductivity of metal oxides but also provide better adsorption of OH in the process of charging–discharging, which can enhance the kinetics of surface reactions [18]. Ferreira et al. [19] synthesized Fe-based spinels with abundant oxygen vacancies by a proteic sol-gel method. Excellent electrochemical performances of CuFe₂O₄ (183 C∙g^‒1 at 0.5 A∙g^‒1 and capacity retention of 98% over 1000 charge–discharge cycles) is attributed to the increased concentration of oxygen vacancies.

Here, oxygen vacancies-engineered NiFe₂O₄ (NiFe₂O_4‒δ) with enhanced conductivity was obtained through a two-step method, including a hydrothermal reaction and a further heat treatment in activated carbon bed. Compared to strategies using reducing agents to introduce oxygen vacancies, this method has many advantages like simplicity, low cost, mass production and environmental friendliness. Compared to the original NiFe₂O₄, the treated NiFe₂O_4‒δ owned a narrower band gap, which was a valid evidence for the enhanced conductivity after heat treatment in activated carbon bed. Besides, the treated NiFe₂O_4‒δ performed a high specific capacitance of 808.02 F∙g^‒1 at the current density of 1 A∙g^‒1, much higher than 219.57 F∙g^‒1 of the original NiFe₂O₄. Moreover, the asymmetric supercapacitor (ASC) composed of NiFe₂O_4‒δ positive electrode and active carbon negative electrode showed an excellent capacity retention of 98.7% after 5000 cycles. All the attractive results prove that the heat treatment process in activated carbon bed is beneficial to improve the conductivity of NiFe₂O₄, which is conducive to the electrochemical performance as a supercapacitor electrode material.

2 Experimental details

2.1 Preparation of NiFe₂O₄

First, 0.891 g NiCl₂·6H₂O, 2.027 g FeCl₃·6H₂O, 0.793 g glucose and 1.5 g urea were put into a beaker. Then 50 mL deionized water was poured into the beaker and mixed onto a magnetic stirring apparatus at 600 r·min⁻¹ for 1 h. Then the well-mixed aqueous solution was poured into an autoclave and heated to 120 °C for 10 h. When the autoclave was naturally cooled to room temperature, the NiFe₂O₄ precursor was washed and collected through suction filtration. Finally, the NiFe₂O₄ was garnered in a muffle furnace at 300 °C for 3 h.

2.2 Preparation of oxygen vacancies-engineered NiFe₂O₄

Oxygen vacancies-engineered NiFe₂O₄ (NiFe₂O_4‒δ) was obtained by a further heat treatment in activated carbon bed. Firstly, 0.5 g obtained NiFe₂O₄ and 0.2 g activated carbon were mixed in an agate mortar through a 30-minute-grinding. Next the well-mixed powders were transferred to an alumina crucible and heated in a tube furnace at 500 °C for 3 h with Ar atmosphere. Finally, NiFe₂O_4‒δ powders were gathered through a heat treatment at 500 °C for 5 h in air. The schematic diagram for synthesis of NiFe₂O_4‒δ is shown in .

Scheme1 The schematic diagram for synthesis of NiFe₂O_4‒δ.

Full size|PPT slide

2.3 Materials characterizations

The surface morphological characterizations and elements distributions of the obtained samples were observed by field emission scanning electron microscope with energy-dispersive spectroscopy (FESEM/EDS, JSM-7800F, JEOL Co., Japan) and field emission transmission electron microscope (TEM, JEM-2100F, JEOL Co., Japan). The phase characterization was analyzed by X-ray diffraction within the scan angle (2θ) range of 10°–90° (XRD, D/max-2500/PC, Rigaku Co., Japan). The chemical states were characterized using X-ray photoelectron spectroscopy (XPS, AXIS Supra, Shimadzu Co., Japan). The optical properties were characterized using a UV-Vis spectrophotometer (UV-Vis, Tu-1901, Purkinje General Co., China). The specific surface areas and pore size distributions were analyzed through N₂ adsorption/desorption isotherms by Brunauer–Emmett–Teller method (ASAP 2460, Micromeritics Co., USA). The functional groups were investigated by Fourier transform infrared spectroscopy (FTIR) in the range of 400–3000 cm^‒1 (Nicolet iS50, Thermo Scientific Co., USA). The oxygen vacancies were analyzed through electron paramagnetic resonance (EPR, A300, Bruker, Germany) in the range of 3200–3800 gauss. The conductivities were investigated by a four-point probe combined with a Keithley 2400C source meter unit.

2.4 Electrochemical measurements

Electrochemical properties of the as-prepared NiFe₂O₄ and NiFe₂O_4‒δ were analyzed by a CHI660E electrochemical station (Shanghai CH Instruments Co., China). In a three-electrode test system, nickel foam was utilized as the current collector of working electrode after 20 minutes’ ultrasonic pretreatment in the order of 2 mol·L⁻¹ HCl, acetone and ethanol, respectively. Electrode materials consisted of 80 wt % obtained NiFe₂O₄ or NiFe₂O_4‒δ, 10 wt % carbon black and 10 wt % polyvinylidene difluoride. After 6 hours’ stirring and dispersing in 1-methyl-2-pyrrolidinone, the obtained mixture was coated on the treated nickel foam uniformly, used as the working electrode. The mass loadings of electrode materials are about 1.5 mg. An 1 cm × 1 cm Pt electrode was utilized as the counter electrode. A standard saturated Ag/AgCl electrode with a salt bridge filled with 3 mol·L⁻¹ KCl aqueous solution was employed to be reference electrode. KOH aqueous solution (1 mol·L⁻¹) was utilized as the electrolyte. Specific capacitance (C) was obtained by Eq. (1) to galvanostatic charge–discharge (GCD) curves:

(1)C=\mathrm{\Delta }Q/\mathrm{\Delta }V={\int }_0^{t}I\mathrm{d}t/(m\mathrm{\Delta }V)=I\mathrm{\Delta }t/(m\mathrm{\Delta }V),

where I is the current (A), Δt is the discharging time (s), m is the coating mass (g), and ΔV is the voltage window (V).

ASCs were also assembled to demonstrate the practical performance of the as-prepared materials. In this device, copper foil was used as the current collector. The positive and negative electrode materials are the mixture of NiFe₂O_4‒δ and activated carbon as 80 wt % materials, respectively. The mass loadings of positive and negative electrodes are confirmed through Eq. (2):

(2)m_{+}{C}_{+}\mathrm{\Delta }{V}_{+}=m_{-}{C}_{-}\mathrm{\Delta }{V}_{-},

where m₊ and m_‒ are mass loadings of positive and negative electrodes, C₊ and C_‒ are specific capacitances of positive and negative electrode materials, and ΔV₊ and ΔV_‒ are potential windows of the two electrodes, respectively.

Energy density (E) and power density (P) were obtained according to Eqs. (3) and (4) on the basis of GCD tests of ASCs:

(3)E=C\mathrm{\Delta }{V}^2/7.2,

(4)P=3600E/\mathrm{\Delta }t,

where ΔV is potential (V), P is power density (W∙kg^‒1), and Δt is the discharging time (s).

3 Results and discussion

Supercapacitor is a kind of energy storage device relying on the contact area between electrode materials’ surfaces and electrolyte accumulating energy [20]. Hence, micromorphology of the materials plays an important role in the energy storage process. Fig.1 shows SEM and TEM images of the NiFe₂O₄ sample synthesized through hydrothermal method and the post-processing NiFe₂O_4‒δ obtained by heat treatment in activated carbon bed. Fig.1(a)–Fig.1(d) show SEM images of NiFe₂O₄ and NiFe₂O_4‒δ. In addition, oxygen vacancies-engineered NiFe₂O₄ treated with different heat treatment temperatures and masses of activated carbon are shown in Fig. S1 (cf. Electronic Supplementary Material, ESM). As seen from Fig.1(a) and Fig.1(b), NiFe₂O₄ synthesized by hydrothermal method grows into nanosheets. The NiFe₂O₄ nanosheets’ lateral sizes are measured to be about 400–600 nm, while the longitudinal sizes are proved to be approximately 5–10 nm. After heat treatment in activated carbon bed, lateral sizes of NiFe₂O_4‒δ reduce obviously, shown in Fig.1(c) and Fig.1(d). About 100–200 nm wide × 5–10 nm thick nanosheets appear after heat treatment. In the heat treatment process, activated carbon reacts with surface oxygen of NiFe₂O₄, which may lead to the surface break-up and form nanosheets with small lateral sizes. Smaller nanosheets possess higher specific surface areas, which can greatly increase the contact area with electrolyte and further enhance the energy storage property [21]. In addition, nanosheets have been proved to perform better than nanoparticles in the application of supercapacitor due to their high specific surface and anti-agglomeration [22]. Lattice structures of NiFe₂O₄ and NiFe₂O_4‒δ are analyzed by TEM. Fig.1(e) and Fig.1(f) show the high-resolution TEM images of NiFe₂O₄ and NiFe₂O_4‒δ. It can be seen that d-spacing of the (311) lattice plane does not change distinctly, which can prove that the crystal structure of NiFe₂O_4‒δ does not change after heat treatment.

Fig.1 SEM images of (a, b) the original NiFe₂O₄ and (c, d) the treated NiFe₂O_4‒δ; TEM images and selected area electron diffraction (SAED) patterns of (e) the original NiFe₂O₄ and (f) the treated NiFe₂O_4‒δ.

Full size|PPT slide

Fig.2 shows some characterizations for phase composition and structure of NiFe₂O₄ and NiFe₂O_4‒δ. In order to verify the components of the original NiFe₂O₄ and the treated NiFe₂O_4‒δ, XRD results are carefully analyzed. Fig.2(a) shows the XRD patterns of the original NiFe₂O₄ and the NiFe₂O_4‒δ. As can be seen, both samples own distinct peaks corresponding to the standard NiFe₂O₄ sample (JCPDS 86-2267). Peaks located in 18.4°, 30.3°, 35.7°, 37.3°, 43.4°, 71.5°, 74.6° and 79.6° can match the diffraction peaks of (111), (220), (311), (222), (400), (620), (533) and (444) crystal planes, respectively. Whereas, compared with the original NiFe₂O₄ sample, NiFe₂O_4‒δ performs better crystallinity. That is because of the higher heat treatment temperature. Meanwhile, an obvious peak of Fe₂O₃ presents in the pattern of original NiFe₂O₄ sample, which appears at 33.2° and is consistent with (104) plane (JCPDS 89-8104). The phenomenon may be caused by the insufficient heat treatment temperature. Here, reference intensity ratio method is used to calculate the phase contents of Fe₂O₃, showing in ESM. Results show that mass ratio of Fe₂O₃ is about 8%. The relatively low content does not have a large effect on the electrochemical performance of original NiFe₂O₄ sample. In addition, influences of heat treatment temperatures and masses of activated carbon on phases are shown in Fig. S2 (cf. ESM).

Fig.2 (a) XRD patterns, (b) FTIR patterns, (c) N₂ adsorption and desorption curves and pore size distributions, (d) EPR spectra, (e) UV-Vis diffuse reflectance spectra, and (f) [F(R)hv]² vs. photon energy plots of the original NiFe₂O₄ and the treated NiFe₂O_4‒δ.

Full size|PPT slide

Fig.2(b) displays FTIR patterns of the original NiFe₂O₄ and the obtained NiFe₂O_4‒δ. Two apparent peaks at the range of 400–600 cm^‒1 appear in both samples. Comparing FTIR characteristic peak data, it can be found that the strong absorption peak at a range of 404–410 cm^‒1 can match the intrinsic stretching vibration of metal atoms in octahedral site (M_oh↔O), while the peak at 589–601 cm^‒1 corresponds to intrinsic stretching vibration of metal atoms in tetrahedral site (M_td↔O) [12,23]. Whereas, comparing the absorption peak positions of the two samples, it can be found that a slightly shift appears after heat treatment with activated carbon, as shown in the inset. Heat treatment with activated carbon can generate a large number of oxygen vacancy defects, which may change the bond length of M↔O, further influence the absorption peak positions [24]. FTIR patterns of the original NiFe₂O₄ and oxygen vacancies-engineered NiFe₂O₄ heat-treated at 500 °C with different masses of activated carbon are shown in Fig. S3 (cf. ESM).

Specific surface area has been proved to be a key factor which may influence the energy storage property of electrode materials [25]. A large specific surface area leads to a fully contact between electrode material and the electrolyte. Hence, N₂ adsorption and desorption measurements of the original NiFe₂O₄ and the obtained NiFe₂O_4‒δ are analyzed. Results show that NiFe₂O_4‒δ performs a larger specific surface area compared with NiFe₂O₄ (48.18 vs. 28.88 m²∙g^‒1), which may be caused by the size reduction of nanosheets during the heat treatment processing with activated carbon. This is consistent with the results obtained from the previous SEM images. As we can see in Fig.2(c), both samples own obvious characteristics of mesoporous materials in the adsorption and desorption curves. Besides, the pore size distribution of the two samples is performed in inset of Fig.2(c). Average pore diameters of the NiFe₂O₄ and NiFe₂O_4‒δ are 18.69 and 19.93 nm, respectively, evidencing that both samples belong to mesoporous materials.

In order to confirm the existence of oxygen vacancies and compare their contents in the original NiFe₂O₄ and the treated NiFe₂O_4‒δ, EPR spectra are exploited. As shown in Fig.2(d), an obvious signal appears at the g value of 2.003 for the two samples, indicating the appearance of oxygen vacancies [26]. In addition, the spectrum of the treated NiFe₂O_4‒δ shows a higher intensity than that of the original sample. This can provide a convincing evidence to prove the increasing of oxygen vacancies in NiFe₂O₄ after heat treatment in activated carbon bed. The mechanism for introducing oxygen vacancies by this method can be divided into two steps. Firstly, activated carbon reacts with oxygen atoms at the surface of NiFe₂O₄, forming surface oxygen vacancies. Next, oxygen atoms in the bulk NiFe₂O₄ diffuse to the surface while surface oxygen vacancies diffuse to the bulk phase, leading to the formation of abundant oxygen vacancies.

Spectral characteristics of the original NiFe₂O₄ and the treated NiFe₂O_4‒δ are studied through UV-Vis diffuse reflectance spectroscopy in the wavelength range of 250–850 nm. As we can see from Fig.2(e), the strong absorption peak appears at wavelengths of around 750 nm for both samples, which is the characteristic peak of NiFe₂O₄ [27]. Optical band gap (E_g) of the two samples are calculated by the Tauc’s equation, as given below:

(5)F(R)=(1-R{)}^2/2R,

(6)(\alpha hv{)}^{1/n}=A\left(hv-E_{\mathrm{g}}\right),

where R is the obtained reflectivity, α is the absorbance coefficient which is proportional to F(R), hv is the light energy, n is a constant depend on the electronic properties of material, and A is a constant. Generally, n = 1/2 is considered as direct transition while n = 2 is considered as indirect transition [28].

The E_g values of NiFe₂O₄ and NiFe₂O_4‒δ are obtained from the intercept on x-axis of the fitted lines of near-straight portion on the [F(R)hv]² vs. photon energy plots, as shown in Fig.2(f). It can be obviously seen that the treated NiFe₂O_4‒δ owns a narrower direct band gap (1.97 eV) compared with the original NiFe₂O₄ (2.07 eV). This proves that the samples after heat treatment with activated carbon have better conductivity, which is beneficial to the energy storage properties [29]. In order to verify the improvement of material conductivity, four-point probe technique is utilized. Results show that the resistance of NiFe₂O_4‒δ is 184.1 MΩ, while an overflow error occurs for NiFe₂O₄ (the maximum resistance value is 200 MΩ). According to the obtained resistance, calculated conductivities of NiFe₂O_4‒δ and NiFe₂O₄ are 6.79 × 10^‒8 and 6.25 × 10^‒8 S∙cm^‒1, respectively. Reason of the increase of conductivity is deemed to be caused by the production of plentiful oxygen vacancies. Researches show that an important conduction mechanism of NiFe₂O₄ can be supposed as the electrons hopping between the same metal elements with different valences, known as the small-polaron hopping model [30,31]. The appearance of oxygen vacancies will inevitably lead to the change of the valence state of metal elements in the sample, leading to the mass production of Fe²⁺/Fe³⁺. The presence of abundant Fe²⁺/Fe³⁺ pairs is believed to benefit the electron conduction of the samples, leading to excellent electrochemical performances.

In order to analyze the element valence distribution of Ni, Fe and O in the original NiFe₂O₄ and the treated NiFe₂O_4‒δ, XPS spectra are obtained, as shown in Fig.3. Fig.3(a) shows the wide-scan XPS patterns of the two samples. It can be seen that typical peaks of Ni, Fe and O appear in both samples, indicating the successfully synthesis of NiFe₂O₄. The molar ratios of Fe/O for the two samples are calculated using Eq. (7):

Fig.3 (a) Wide-scan XPS patterns; deconvolutions of (b) Ni 2p, (c) Fe 2p and (d) O 1s for the original NiFe₂O₄ and the treated NiFe₂O_4‒δ.

Full size|PPT slide

(7)n_1/n_2=\left(I_1/S_1\right)/\left(I_2/S_2\right),

where n₁/n₂ is the molar ratio, I is the elemental peak area integration, and S is the sensitivity factor.

Results show that molar ratio of Fe/O in the original NiFe₂O₄ is 0.531, smaller than 0.589 of the treated NiFe₂O_4‒δ. This can demonstrate the reduction of O content in the treated sample, that is, the increase of oxygen vacancies.

The distributions and valance states of Ni, Fe and O in the two samples are analyzed by deconvolutions of these peaks utilizing Gaussian fitting. Fig.3(b)–3(d) detail the deconvolutions of Ni, Fe and O, respectively. As shown in Fig.3(b), two peaks located at binding energy of 854.7 and 856.0 eV form the Ni 2p_3/2, representing Ni²⁺ in the octahedral voids and tetrahedral voids, respectively. Moreover, the corresponding peaks at 872.2 and 873.5 eV compose the Ni 2p_1/2. In addition, two peaks at 861.4 and 879.2 eV are considered as satellite peaks of Ni 2p_3/2 and Ni 2p_1/2, respectively [32].

From Fig.3(c), the peak of Fe 2p consists of two peaks of Fe 2p_1/2 and 2p_3/2, both of which can be deconvoluted into three peaks considered as Fe²⁺, Fe³⁺_oct and Fe³⁺_tet. Specifically, peaks of 709.9 (Fe²⁺), 711.0 (Fe³⁺_oct) and 712.4 (Fe³⁺_tet) eV constitute Fe 2p_3/2 spectra, while the curve of Fe 2p_1/2 is deconvoluted into peaks at 723.2 (Fe²⁺), 724.3 (Fe³⁺_oct) and 725.7 (Fe³⁺_tet) eV [33]. Molar ratios of Fe²⁺/Fe³⁺ in the two samples are calculated through the integral areas of these deconvolution peaks. As a result, molar ratio of 28.8% (Fe²⁺/Fe³⁺) in the treated NiFe₂O_4‒δ is higher than that in the original NiFe₂O₄ as 22.0%, which can prove that partial Fe³⁺ is reduced to Fe²⁺. Meanwhile, this explains the increase of Fe²⁺/Fe³⁺ electron pairs in the treated sample, which is beneficial to the conductivity and further to the electrochemical performances.

Fig.3(d) shows the O 1s spectra of the two samples and their deconvolutions. Peaks located at 529.9, 531.6 and 533.2 eV are identified as the lattice oxygen (O_L), oxygen vacancy (O_V) and oxygen in physical or chemical absorption H₂O (O_OH) [34]. Additionally, the change of oxygen vacancy contents in the two samples is determined according to the deconvoluted integrated area of O_V. The O_V peak area percentage of NiFe₂O_4‒δ is calculated to be 25%, higher than that of NiFe₂O₄ (23.5%). This result provides another evidence of the increasing of oxygen vacancy in NiFe₂O₄ after heat treatment in activated carbon bed.

In order to figure out the electrochemical properties of the prepared NiFe₂O₄ and NiFe₂O_4‒δ, a three-electrode system was utilized. Results of the electrochemical performances are shown in Fig.4. Cyclic voltammetry (CV) curses of NiFe₂O₄ and NiFe₂O_4‒δ are shown in Fig.4(a) and 4(b), respectively. Two apparent specific redox peaks can be observed in every curve for both samples, which represents typical battery-like characteristics [35]. In general, the contribution of the redox peaks is considered to be the following electrochemical redox reaction [13]:

Fig.4 CV curves of (a) NiFe₂O₄ and (b) NiFe₂O_4‒δ; (c) plot of lg(v) versus lg(i) for NiFe₂O₄ and NiFe₂O_4‒δ; contributions of different electrochemical behaviors in the energy storage processes for (d) NiFe₂O₄ and (e) NiFe₂O_4‒δ; (f) plot of J_p versus v^1/2 for NiFe₂O₄ and NiFe₂O_4‒δ; GCD curves of (g) NiFe₂O₄ and (h) NiFe₂O_4‒δ; (i) Nyquist plots of NiFe₂O₄ and NiFe₂O_4‒δ.

Full size|PPT slide

{\mathrm{N}\mathrm{i}\mathrm{F}\mathrm{e}}_2{\mathrm{O}}_4+{\mathrm{H}}_2\mathrm{O}+{\mathrm{O}\mathrm{H}}^{-1}\leftrightarrow \mathrm{N}\mathrm{i}\mathrm{O}\mathrm{O}\mathrm{H}+2\mathrm{F}\mathrm{e}\mathrm{O}\mathrm{O}\mathrm{H}+{\mathrm{e}}^{-1}

Additionally, the integral areas of CV curves can roughly demonstrate the electrochemical storage properties [36]. From Fig.4(a) and 4(b), the CV curves of the sample after heat treatment with activated carbon show higher integral areas than the original NiFe₂O₄, indicating a higher energy storage performance. Besides, electrochemical reaction kinetics analysis of the two samples can be validly exposed according to CV curves. Generally speaking, the relationship of peak current and scan rate can be expressed by Eq. (8):

(8)i_{\mathrm{p}}=a{v}^b,

where i_p is the peak current (A), v is the scan rate (mV∙s^‒1), and a and b are variable parameters. Commonly, b = 1 represents capacitance behavior control while b = 0.5 implies diffusion behavior control [37]. Fig.4(c) shows the plots of lg(v) vs. lg(i) for the two samples. In this chart, b values are obtained in accord with the slopes of the fitted lines. As a result, b values for both samples are calculated to be about 0.6, which proves that electrochemical activities for both samples are mainly determined by diffusion behavior.

Equation (8) can also be deformed into the following formula:

(9)i_{\mathrm{p}}=k_1{v}^{0.5}+k_{2}v,

where k₁ and k₂ are variable parameters. In Eq. (9), peak current (i_p) is divided into two parts. The k₁v^0.5 (I_D) represents the contribution of diffusion behavior, while the k₂v (I_C) owes to the capacitive behavior [38]. Specifically, the contributions of different electrochemical behaviors in the energy storage processes for the original NiFe₂O₄ and the treated NiFe₂O_4‒δ at different scan rates are shown in Fig.4(d) and Fig.4(e), respectively. For both samples, the contributions of diffusion processes gradually decrease while capacitive processes continuously enlarge as the scan rates increase. At a low scan rate, the ions diffusion behavior between electrode and electrolyte can take place to a large extent. Hence, diffusion processes occupy a major proportion in the energy storage processes. While as the scan rate increase, ions diffusions with low reaction rates cannot match the high scan rates, resulting in inadequate diffusion reactions. Thus, increments of capacitive processes happen naturally [39].

Furthermore, data obtained from the CV curves can also be utilized to calculate the average ion diffusion coefficient of OH according to the Randles–Sevcik equation, as shown in Eq. (10):

(10)J_{\mathrm{p}}/{\nu }^{1/2}=0.4463n^{3/2}{F}^{3/2}{CSR}^{-1/2}{T}^{-1/2}{D}^{1/2},

where J_p is the peak current density of the CV curve (A∙g^‒1), v is the scan rate (V∙s^‒1), n is the number of transferred electrons, F is the Faraday constant (C∙mol^‒1), C is the OH concentration of the electrolyte (mol∙cm^‒3), S is the specific surface area (cm²∙g^‒1), T is the temperature (K), and D is the ion diffusion coefficient (cm²∙s^‒1).

Fig.4(f) shows the fitted lines of the peak current densities (J_p) versus square of scan rates (v^1/2) for the original NiFe₂O₄ and the treated NiFe₂O_4‒δ. In accordance with Eq. (10), the ion diffusion coefficients of OH for NiFe₂O₄ and NiFe₂O_4‒δ are 8.25 × 10^‒11 and 2.41 × 10^‒10 cm²∙s^‒1, respectively. The greater ion diffusion coefficient of NiFe₂O_4‒δ represents a faster ion diffusion rates during electrochemical reactions, which is beneficial to the electrochemical properties.

GCD curves at different current densities for NiFe₂O₄ and NiFe₂O_4‒δ are shown in Fig.4(g) and Fig.4(h), respectively. Apparent charging and discharging platforms can be observed in both samples, which prove that the two materials behave as classic battery-like electrode materials [40]. Specific capacitances of the two samples are calculated using Eq. (1). As a result, NiFe₂O_4‒δ performs high specific capacitances of 808.02, 751.6, 644.77, 609.76 and 537.78 F∙g^‒1 at the current densities of 1, 2, 5, 8 and 10 A·g^‒1, respectively.

In comparison, specific capacitances of the original NiFe₂O₄ are slightly poor, which were calculated to be 219.57, 210.68, 196.66, 181.33 and 115.56 F·g^‒1. Tab.1 shows the electrochemical performance comparation with some previous works. Moreover, NiFe₂O_4‒δ possesses a better capacitance retention than the original NiFe₂O₄. At the current density of 10 A·g^‒1, the capacitance of NiFe₂O_4‒δ can retain about 67%, while the original NiFe₂O₄ has only 53% left. Besides, specific capacitances of the oxygen vacancies-engineered NiFe₂O₄ heat-treated at 500 °C with different masses of activated carbon are calculated according to GCD curves, showing in Fig. S4 (cf. ESM).

Tab.1 Specific capacitances of the previous works on NiFe₂O₄ supercapacitors

Material	Specific capacitance/(F·g‒1)	Current density	Electrolyte	Ref.
Beehive-like NiFe2O4	560	1 A·g‒1	1 mol·L−1 KOH	[41]
2D porous layered NiFe2O4	425.62	1 A·g‒1	6 mol·L−1 KOH	[42]
NiFe2O4	466	1 A·g‒1	1 mol·L−1 H2SO4	[43]
NiFe2O4/rGO	301	0.5 mA·cm‒1	1 mol·L−1 KOH	[44]
NiFe2O4	219.57	1 A·g‒1	1 mol·L−1 KOH	This work
Oxygen vacancies engineered NiFe2O4‒δ	808.02	1 A·g‒1	1 mol·L−1 KOH	This work

Fig.4(i) displays the results of electrochemical impedance spectroscopy tests of NiFe₂O₄ and NiFe₂O_4‒δ and an equivalent circuit model used for fitting processing. Thereinto, R_s and R_ct represent equivalent series resistances and charge transfer resistances, respectively, obtained from the fitting of high frequency area semicircles. Results indicate that R_s values of NiFe₂O₄ and NiFe₂O_4‒δ are 0.99 and 0.94 Ω. Meanwhile, NiFe₂O_4‒δ shows a lower R_ct value of 21.07 Ω than that of NiFe₂O₄, which is 26.67 Ω. This represents that the conductivity between the electrode material and the electrolyte is higher in the electrochemical system of NiFe₂O_4‒δ.

To confirm the application prospect of the obtained NiFe₂O_4‒δ, an ASC composed of a NiFe₂O_4‒δ positive electrode and an active carbon negative electrode was fabricated. Results of electrochemical performances of the fabricated ASC are shown in Fig.5. Fig.5(a) displays CV curves of the obtained ASC at different scan rates. As we can see, a high 0–1.5 voltage window is achieved in this device. This is powerful evidence that the obtained NiFe₂O_4‒δ has good application potentials. To further judge the energy storage properties of the assembled device, GCD curves were obtained through chronopotentiometry method, showing in Fig.5(b). Calculated by Eq. (1), specific capacitances of the device at current densities of 0.5, 1, 2, 5 and 10 A·g^‒1 are attained to be 56.7, 45.3, 36.0, 30.0 and 26.0 F·g^‒1. The slight polarization in CV curve as well as voltage drop in GCD curve are caused by the intercalation and de-intercalation of OH ions in a solid phase structure [45,46]. Fig.5(c) shows Ragone plots of the obtained NiFe₂O_4‒δ//AC device and other researchers’ works [47–50]. It can be seen that at a power density of 375 W·kg^‒1, the energy density of the obtained NiFe₂O_4‒δ//AC device can reach 17.7 Wh·kg^‒1, demonstrating good application performance. In addition, Fig.5(d) displays the cycling stability and coulombic efficiency at 3 A·g^‒1. As seen, the device can maintain 98.7% capacitance after 5000 cycles with a high coulombic efficiency of 99.4%, preforming excellent reversibility and cyclic stability. Besides, Fig.5(e) and 5(f) show the SEM images of the positive electrode before and after cycling. It can be seen that NiFe₂O_4‒δ nanosheets and carbon particles are evenly mixed. After 5000 cycles, the morphologies of NiFe₂O_4‒δ nanosheets hardly changed, indicating the excellent cyclic stability. All these attractive results indicate that NiFe₂O_4‒δ//AC ASC device has good electrochemical performance and enormous application potential.

Fig.5 (a) CV curves, (b) GCD curves, (c) Ragone plots, (d) cycling stability curve and coulombic efficiency curve of the NiFe₂O_4‒δ//AC ASC; SEM images of the positive electrode (e) before and (f) after cycling.

Full size|PPT slide

4 Conclusions

In summary, NiFe₂O_4‒δ with abundant oxygen vacancies was obtained through a further heat treatment in activated carbon bed on the basis of hydrothermal-synthesized NiFe₂O₄. SEM results showed that morphologies of the samples were nanosheets, moreover, lateral sizes of the NiFe₂O_4‒δ after further heat treatment reduced obviously. The smaller size led to a higher specific surface area. Oxygen vacancies can enhance the electrical conductivities of metal oxides. From the results of UV-Vis diffuse reflectance spectroscopy, it could be obviously seen that the treated NiFe₂O_4‒δ owned a narrower band gap (1.97 eV) compared with the original NiFe₂O₄ (2.07 eV). That was valid evidence for the enhanced conductivity after heat treatment in activated carbon bed. In addition, oxygen vacancies can tune the cation adsorption Gibbs free energy and supply a cation transfer path and an extra adsorption site. Further, the treated NiFe₂O_4‒δ performed a high specific capacitance of 808.02 F·g^‒1 at the current density of 1 A·g^‒1, much higher than 219.57 F·g^‒1 of the original NiFe₂O₄. Moreover, the ASC of NiFe₂O_4‒δ//active carbon displayed a high energy density of 17.7 Wh·kg^‒1 at a power density of 375 W·kg^‒1. To sum up, this work provides an effective way to improve the conductivity of metal oxides, which is beneficial to the application of metal oxides in the field of supercapacitors.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

This work was supported by Major Basic Research Projects of Shandong Natural Science Foundation (Grant No. ZR2018ZB0104), Science and Technology Development Project of Shandong Province (Grant Nos. 2016GGX102003 and 2017GGX20105), and Natural Science Foundation of Shandong Province (Grant No. ZR2017BEM032).

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://doi.org/10.1007/s11705-023-2352-6 and is accessible for authorized users.

References

1 Lama Tamang T, Mohamed S G, Dhakal G, Shim J J. Morphology controlling of manganese-cobalt-sulfide nanoflake arrays using polyvinylpyrrolidone capping agent to enhance the performance of hybrid supercapacitors. Journal of Colloid and Interface Science, 2022, 624: 494–504

2 Wang Q, Qu Z, Chen S, Zhang D. Metal organic framework derived P-doping CoS@C with sulfide defect to boost high-performance asymmetric supercapacitors. Journal of Colloid and Interface Science, 2022, 624: 385–393

3 Lv H, Xiao Z, Zhai S, Hao J, Tong Y, Wang G, An Q. Construction of nickel ferrite nanoparticle-loaded on carboxymethyl cellulose-derived porous carbon for efficient pseudocapacitive energy storage. Journal of Colloid and Interface Science, 2022, 622: 327–335

4 Dharmasiri B, Stanfield M K, Randall J D, Usman K A S, Qin S A, Razal J M, Doeven E H, Francis P S, Eyckens D J, Yin Y, Andersson G G, Henderson L C. Multifunctional polymeric surface coatings of carbon fibre electrodes for enhanced energy storage performance. Chemical Engineering Journal, 2022, 447: 137560

5 Pang H, Wang M, Sun P, Zhang W, Wang D, Zhang R, Qiao L, Wang W, Gao M, Li Y, Chen J, Liang K, Kong B. Super-assembled compressible carbon frameworks featuring enriched heteroatom defect sites for flexible Zn-air batteries. NPG Asia Materials, 2023, 15(1): 15

6 Zhang H, Wang J, Duan H, Ren J, Zhao H, Zhou C, Qi J. Mn³⁺ partially substituting the Ni³⁺ of NiCo₂O₄ enhance the charge transfer kinetics and reaction activity for hybrid supercapacitor. Applied Surface Science, 2022, 597: 153617

7 Li Y, Zhu G, Xu X, Chen L, Lu T, Hill J P, Pan L, Yamauchi Y. Embedding metal-organic frameworks for the design of flexible hybrid supercapacitors by electrospinning: synthesis of highly graphitized carbon nanofibers containing metal oxide nanoparticles. Small Structures, 2022, 3(9): 2200015

8 Liu J, Wang Z, Liu Q, Li S, Wang D, Zheng Z. Rational design of freestanding and high-performance thick electrode from carbon foam modified with polypyrrole/polydopamine for supercapacitors. Chemical Engineering Journal, 2022, 447: 137562

9 Chen J, Liu B, Cai H, Liu S, Yamauchi Y, Jun S C. Covalently interlayer-confined organic-inorganic heterostructures for aqueous potassium ion supercapacitors. Small, 2023, 19(4): 2204275

10 Zhou Y, Wei L, Li C, Han Y, Xu J, Jia Z, Sun J, Chen H, Song Y, Ouyang X, Wang X, Zhu J, Fu Y. Nanostructure and phase engineering integration of amorphous Ni-Co sulfide/crystalline MnS/rGO cathode and ultra-small Fe₂O₃ nanodots/rGO anode for all-solid-state asymmetric supercapacitors. Journal of Energy Storage, 2022, 45: 103765

11 Guo M, Sun J, Liu Y, Huangfu C, Wang R, Han C, Qu Z, Wang N, Zhao L, Zheng Q. Optimizing Fe₂O₃-based supercapacitor cathode with tunable surface pseudocapacitance via facile in situ vulcanization process. Journal of Electroanalytical Chemistry, 2021, 901: 115785

12 Bandgar S B, Vadiyar M M, Ling Y C, Chang J Y, Han S H, Ghule A V, Kolekar S S. Metal precursor dependent synthesis of NiFe₂O₄ thin films for high-performance flexible symmetric supercapacitor. ACS Applied Energy Materials, 2018, 1(2): 638–648

13 Mordina B, Kumar R, Neeraj N S, Srivastava A K, Setua D K, Sharma A. Binder free high performance hybrid supercapacitor device based on nickel ferrite nanoparticles. Journal of Energy Storage, 2020, 31: 101677

14 Yu Z Y, Chen L F, Yu S H. Growth of NiFe₂O₄ nanoparticles on carbon cloth for high performance flexible supercapacitors. Journal of Materials Chemistry A: Materials for Energy and Sustainability, 2014, 2(28): 10889–10894

15 Schmidt R, Basu A, Brinkman A W, Klusek Z, Datta P K. Electron-hopping modes in NiMn₂O_4+δ materials. Applied Physics Letters, 2005, 86(7): 073501

16 Yoon S J, Lee S H, Kim K H, Ahn K S. Electrical and magnetic properties of spinel ZnCr_2−xFe_xO₄ (0 ≤ x ≤ 1.0). Materials Chemistry and Physics, 2002, 73(2–3): 330–334

17 Zhang A, Gao R, Hu L, Zang X, Yang R, Wang S, Yao S, Yang Z, Hao H, Yan Y M. Rich bulk oxygen vacancies-engineered MnO₂ with enhanced charge transfer kinetics for supercapacitor. Chemical Engineering Journal, 2021, 417: 129186

18 Xu L, Pan G, Yu C, Li J, Gong Z, Lu T, Pan L. Co-doped MnO₂ with abundant oxygen vacancies as a cathode for superior aqueous magnesium ion storage. Inorganic Chemistry Frontiers, 2023, 10(6): 1748–1757

19 Ferreira L S, Silva T R, Silva V D, Simões T A, Araújo A J M, Morales M A, Macedo D A. Proteic sol-gel synthesis, structure and battery-type behavior of Fe-based spinels (MFe₂O₄, M = Cu, Co, Ni). Advanced Powder Technology, 2020, 31(2): 604–613

20 Zhang K, Zeng H Y, Li H B, Xu S, Lv S B, Wang M X. Controllable preparation of CuCo₂S₄ nanotube arrays for high-performance hybrid supercapacitors. Electrochimica Acta, 2022, 404: 139681

21 Gu J, Fan X, Liu X, Li S, Wang Z, Tang S, Yuan D. Mesoporous manganese oxide with large specific surface area for high-performance asymmetric supercapacitor with enhanced cycling stability. Chemical Engineering Journal, 2017, 324: 35–43

22 Gao X, Wang W, Bi J, Chen Y, Hao X, Sun X, Zhang J. Morphology-controllable preparation of NiFe₂O₄ as high performance electrode material for supercapacitor. Electrochimica Acta, 2019, 296: 181–189

23 Carpenter G, Sen R, Malviya N, Gupta N. Microwave-assisted synthesis and characterization of nickel ferrite nanoparticles. AIP Conference Proceedings, 2015, 1675: 020029

24 Stella C, Prabhakar D, Prabhu M, Soundararajan N, Ramachandran K. Oxygen vacancies induced room temperature ferromagnetism and gas sensing properties of Co-doped TiO₂ nanoparticles. Journal of Materials Science Materials in Electronics, 2016, 27(2): 1636–1644

25 Wang C, Sui G, Guo D, Li J, Zhuang Y, Guo W, Zhou Y, Yang X, Chai D F. Inverted design of oxygen vacancies modulated NiCo₂O₄ and Co₃O₄ microspheres with superior specific surface area as competitive bifunctional materials for supercapacitor and hydrogen evolution reaction. Journal of Energy Storage, 2022, 49: 104083

26 Wang C, Sui G, Guo D, Li J, Ma X, Zhuang Y, Chai D F. Oxygen vacancies-rich NiCo₂O_4‒4x nanowires assembled on porous carbon derived from cigarette ash: a competitive candidate for hydrogen evolution reaction and supercapacitor. Journal of Energy Storage, 2022, 50: 104280

27 Sharifi S, Yazdani A, Rahimi K. Effect of Co²⁺ content on supercapacitance properties of hydrothermally synthesized Ni_1‒xCo_xFe₂O₄ nanoparticles. Materials Science in Semiconductor Processing, 2020, 108: 104902

28 Boukhemikhem Z, Brahimi R, Rekhila G, Fortas G, Boudjellal L, Trari M. The photocatalytic hydrogen formation and NO₂⁻ oxidation on the hetero-junction Ag/NiFe₂O₄ prepared by chemical route. Renewable Energy, 2020, 145: 2615–2620

29 Aafiya M, Abushad M, Arshad S, Naseem H, Ahmed A, Ansari V K, Chakradhary S, Husain W. Synthesis and role of structural disorder on the optical, magnetic and dielectric properties of Zn doped NiFe₂O₄ nanoferrites. Journal of Molecular Structure, 2022, 1253: 132205

30 He L, Ling Z. Studies of temperature dependent AC impedance of a negative temperature coefficient Mn-Co-Ni-O thin film thermistor. Applied Physics Letters, 2011, 98(24): 242112

31 Nayak P, Nayak S K, Satpathy B. Structural, electro-chemical and conduction mechanism in spinel NiFe₂O₄/NFO supercapacitor electrode material. Materials Science in Semiconductor Processing, 2022, 143: 106543

32 Gao X, Bi J, Gao J, Meng L, Xie L, Liu C. Partial sulfur doping induced lattice expansion of NiFe₂O₄ with enhanced electrochemical capacity for supercapacitor application. Electrochimica Acta, 2022, 426: 140739

33 Munonde T S, Zheng H, Matseke M S, Nomngongo P N, Wang Y, Tsiakaras P. A green approach for enhancing the electrocatalytic activity and stability of NiFe₂O₄/CB nanospheres towards hydrogen production. Renewable Energy, 2020, 154: 704–714

34 Ma Q, Cui F, Zhang J, Qi X, Cui T. Surface engineering of Co₃O₄ nanoribbons forming abundant oxygen-vacancy for advanced supercapacitor. Applied Surface Science, 2022, 578: 152001

35 Wang D G, Liang Z, Gao S, Qu C, Zou R. Metal-organic framework-based materials for hybrid supercapacitor application. Coordination Chemistry Reviews, 2020, 404: 213093

36 Nagarani S, Sasikala G, Satheesh K, Yuvaraj M, Jayavel R. Synthesis and characterization of binary transition metal oxide/reduced graphene oxide nanocomposites and its enhanced electrochemical properties for supercapacitor applications. Journal of Materials Science Materials in Electronics, 2018, 29(14): 11738–11748

37 Yang P, Wu Z, Jiang Y, Pan Z, Tian W, Jiang L, Hu L. Fractal (Ni_xCo_1−x)₉Se₈ nanodendrite arrays with highly exposed (011) surface for wearable, all-solid-state supercapacitor. Advanced Energy Materials, 2018, 8(26): 1801392

38 Jiang J, Li Z, He X, Hu Y, Li F, Huang P, Wang C. Novel skutterudite CoP₃ based asymmetric supercapacitor with super high energy density. Small, 2020, 16(31): 2000180

39 Xie M, Zhou M, Zhang Y, Du C, Chen J, Wan L. Freestanding trimetallic Fe-Co-Ni phosphide nanosheet arrays as an advanced electrode for high-performance asymmetric supercapacitors. Journal of Colloid and Interface Science, 2022, 608: 79–89

40 Shang Y, Ma S, Wei Y, Yang H, Xu Z. Flower-like ternary metal of Ni-Co-Mn hydroxide combined with carbon nanotube for supercapacitor. Ionics, 2020, 26(7): 3609–3619

41 Samuel E, Aldalbahi A, El-Newehy M, El-Hamshary H, Yoon S S. Nickel ferrite beehive-like nanosheets for binder-free and high-energy-storage supercapacitor electrodes. Journal of Alloys and Compounds, 2021, 852: 156929

42 Huang T, Cui W, Qiu Z, Hu Z, Zhang Z. 2D porous layered NiFe₂O₄ by a facile hydrothermal method for asymmetric supercapacitor. Ionics, 2021, 27(3): 1347–1355

43 Malarvizhi M, Meyvel S, Sandhiya M, Sathish M, Dakshana M, Sathya P, Thillaikkarasi D, Karthikeyan S. Design and fabrication of cobalt and nickel ferrites based flexible electrodes for high-performance energy storage applications. Inorganic Chemistry Communications, 2021, 123: 108344

44 Askari M B, Salarizadeh P. Binary nickel ferrite oxide (NiFe₂O₄) nanoparticles coated on reduced graphene oxide as stable and high-performance asymmetric supercapacitor electrode material. International Journal of Hydrogen Energy, 2020, 45(51): 27482–27491

45 Wei Y, Zou X, Cen C, Zhang B, Xiang B, Hao J, Wang B, Deng M, Hu Q, Wei S. Controlling the electrochemical activity of dahlia-like β-NiS@rGO by interface polarization. Dalton Transactions, 2023, 52(5): 1345–1356

46 Zhang M, Chen Y, Yang D, Li J. High performance MnO₂ supercapacitor material prepared by modified electrodeposition method with different electrodeposition voltages. Journal of Energy Storage, 2020, 29: 101363

47 Zhang S, Yin B, Wang Z, Peter F. Super long-life all solid-state asymmetric supercapacitor based on NiO nanosheets and α-Fe₂O₃ nanorods. Chemical Engineering Journal, 2016, 306: 193–203

48 Cai W, Lai T, Dai W, Ye J. A facile approach to fabricate flexible all-solid-state supercapacitors based on MnFe₂O₄/graphene hybrids. Journal of Power Sources, 2014, 255: 170–178

49 Yang S, Han Z, Sun J, Yang X, Hu X, Li C, Cao B. Controllable ZnFe₂O₄/reduced graphene oxide hybrid for high-performance supercapacitor electrode. Electrochimica Acta, 2018, 268: 20–26

50 Sethi M, Shenoy U S, Bhat D K. A porous graphene-NiFe₂O₄ nanocomposite with high electrochemical performance and high cycling stability for energy storage applications. Nanoscale Advances, 2020, 2(9): 4229–4241